TFI - CADTips Challenge
So far it seems to be a hit with everyone. Most people getting it just slightly wrong, handful of people got it right so far!
I'm still not 100% sure how you would do this in Fusion, it's missing the feature which plays a big part in the formation of this design so I'm not sure how you'd approach this... but super respect will go out to anyone who manages to nail this in Fusion! @lars_christensen nudge nudge be a lamb and smash this one out the park for Fusion!
@Neil_Cross @TheCADWhisperer In Fusion 360 this is all hand modeling the final form.
I don't work with Inventor but I assume it has a bend function that allows to deform and flat modeled object.
I can do that in my other non Autodesk CAD software as well, but in Fusion 360 that does not exist.
I've not come to conclusion yet if that can be done accurately, but I'd have to use that clunky web viewer to take measurements and already could not figure out how to get that to work. LOL.
Maybe after a good nights sleep 😉
I found a few workflows, some quicker than others but the final problem is the lack of the ability to extrude from a double convex shape or bend a flat shape. I am sure you saw this right away but I had to try things. This angle should be 90 deg...
Hi Peter - I wouldn't confirm or deny if that's right or not as I wouldn't want the correct answer to be available for everyone to see! Part of the mystery of the challenge is that people don't know if they've done it right ;0)
Here's mine. I doubt that it, or anything done in Fusion 360, will match the Inventor specs unless the method/formula Inventor uses to distort the flat model into the compound curve model is known.
Here's Detail C:
ETFrench
The flat pattern state is an impossible state. If the inner lip vertical surfaces are normal to outer lip's top surface, then its top and bottom widths must be different. There is also no reference surface given. If it's the top surface of the inner lip, then the outer lip cannot be 9mm wide and still be 100mm x 130mm.
My Detail C used the top surface of the outer lip as the reference surface. The dimension shown (9.0000mm) is point to point while it should be the length of the arc segment which is currently 34 nanometers longer than it should be 😁
ETFrench
Well 21 people so far have managed to get it done correctly using the original detailing!
FYI I dunno if this is what you meant, but the flat state isn't a flat pattern in the context of sheet metal. The views and dims exist purely to give someone enough info to remodel it, and also a clue as to what the most efficient way is to get there, which it definitely does as a good few people have managed it!
So I modeled this in ZW3D with a method I assume much, much closer to the one that is likely used in Inventor.
Needless to say it only took a small fraction of the time and was straight forward.
I exported as .step, imported into Fusion 360, applied the SS material and took the properties. I am off by 2 milligrams on the weight compared to my Fusion 360 design. Close enough in my book, because:
1. We are exporting/importing NURBS surfaces and the algorithms in ZW3D unlikely are to be exactly the same as the one in Inventor.
2. The Fusion 360 design might not be milligram comparable to the one from Inventor due to the fully manual modeling approach required by Fusion 360. While both use the same geometric modeling kernel I would not be surprised to see some difference.
Edit:
I did another little experiment. I removed the chamfer of the model imported from ZW3D and re-applied a Fusion 360 native chamfer of the same size. That alone caused a drop in weight by another milligram.
@TrippyLighting wrote:Edit:
I did another little experiment. I removed the chamfer of the model imported from ZW3D and re-applied a Fusion 360 native chamfer of the same size. That alone caused a drop in weight by another milligram.
I did similar experiment in entirely in Inventor.
I removed the chamfer from flat state and then added it back in the double bent state and got a slight difference in mass properties. Superimposing one part over the other shows slight difference in boundary edge (at least visual, I did not bother to attempt measurement of the observed difference).
Section view shown below.
Given that the contest shows the Chamfer in the flat state before "bending", I would say that is when it should be placed, but this presents a conundrum for modeling in software that does not have a Bend feature.
I would say the Bend feature itself is a confounding variable as I don't think it matches real world condition of pressing the prescribed flat into the double curvature. I think the part would have to be molded in finished state to match whatever the Bend algorithm produces.
Edit: Given this demonstration of the Chamfer feature, all of the other features would also be subject to the same potential differences in attempting to model in finished state vs "Bending" from the given flat state.
I modeled the part in both Inventor and in SolidWorks.
The Flat Form mass properties are identical in both.
If this was representative of a real world - take this flat part and bend as shown, the Mass and Volume would be identical in both flat and bent form.
But purely as a modeling challenge the Volume of material does change.
My results are significantly different between SolidWorks and Inventor and bent and flat (in each) even though Flat Form was identical.
I will post videos of my technique in both programs for others to verify if I did or did not use equivalent techniques in both SolidWorks and Inventor and replicate my results.
